Solutions · Sectors
Value bettors
A model you can check, not a price you have to take on trust.
Most tips circulate with no visible method: a number, a confident tone, no trace of how it was arrived at. There is no way, on that basis, to judge whether the figure is sound or merely well presented.
A model distinct from the one on this site
The Dixon-Coles model published here remains our reference for editorial content: it is documented because it has to be explainable and contestable. The model that identifies pricing gaps is a different one, proprietary, calibrated for a different task — performance against market odds, not the teaching of a result. Two aims call for two models.
Where it is used
In the Docteur Bet app, which flags the gaps between the computed probability and the price on offer. For professional sports traders, the raw probabilities from that same model are available through an API, to be integrated into your own trading or portfolio tools.
What you can verify
The model’s full result history is available — not only the wins that get highlighted. The method is explained in outline: what the model takes into account, what it does not capture, and where its limits are. The detail of its implementation is not published, unlike Dixon-Coles: that is what carries the product’s value. Its performance, however, stays verifiable over time.
What it changes in practice
A bet placed because the computation shows a measured gap in value, not because an account with a large following recommended it. And a limit stated up front: a statistical edge is not a certainty, and we write that as plainly in the app as on this page.
A few concrete examples
- The value gaps identified, in the app
- The model’s full history, losses included
- Raw probabilities through the API, for professional use